Point Light Source Detection for Accurate HDR Image Composition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting point light sources in images, such as those used in high-dynamic range imaging, are inaccurate due to the presence of point light sources or white objects, leading to difficulties in generating HDR images with better definition.
Innovation Solution
A method that filters images based on low-brightness and high-brightness pixels to determine if they meet dark environment and point light source size requirements, using pixel proportions and average brightness thresholds to accurately identify point light sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If pixel brightness threshold detection is used to identify point light sources, then the detection process is simple, but the detection accuracy is low due to false positives from white objects
Solution Approach 1:
The patent segments the detection process into multiple independent stages: dark environment judgment based on low-brightness pixel proportion, point light source region extraction based on high-brightness pixels, and size verification. This segmentation allows each stage to focus on specific characteristics, improving overall detection accuracy while maintaining manageable complexity
Solution Approach 2:
The patent introduces multiple parameters beyond simple brightness thresholds, including low-brightness pixel proportion (dark environment parameter), high-brightness pixel count (point light source parameter), and region area (size parameter). By changing from a single brightness parameter to multiple parameters, the system achieves more accurate detection
2Measurement precision
If multiple detection parameters are introduced to improve accuracy, then detection precision improves, but the detection process becomes more complex
Solution Approach 1:
The patent divides the complex multi-parameter detection into sequential simple judgments: first check dark environment using low-brightness pixel proportion, then extract high-brightness regions, finally verify point light source size. This segmentation makes the complex process easier to implement and understand
Solution Approach 2:
The patent performs preliminary classification by first judging whether the image meets dark environment requirements before proceeding to point light source detection. This preliminary action filters out unsuitable images early, simplifying subsequent detection processes
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
This application relates to the field of image processing, and provides a point light source image detection method and an electronic device. The method includes: obtaining to-be-tested images, and determining pixel brightnesses of the to-be-tested images; determining low-brightness pixels and high-brightness pixels in the images based on the pixel brightnesses; filtering, based on the low-brightness pixels, out an image that does not meet a preset dark environment requirement, and filtering, based on the high-brightness pixels, out an image that does not meet a preset point light source size; and determining, based on a result of the filtering, whether a to-be-tested image is a point light source image. Pixels in images are classified into low-brightness pixels and high-brightness pixels. For the low-brightness pixels, the images are screened based on a preset dark environment requirement, so that the selected point light source images meet the preset dark environment requirement. Moreover, the point light source images are screened based on a size characteristic of a point light source, to avoid repetition with an overexposed region in a HDR detection process, thereby improving precision of point light source image detection.